OBJECTIVETo investigate whether the risk for autoimmune diabetes in adults differs between socioeconomic groups and to compare such risk with that for type 2 diabetes.RESEARCH DESIGN AND METHODSThe inhabitants of the Norwegian county of Nord-Trøndelag were investigated by questionnaires and clinical examinations on three occasions during 1984–2008. We used information from a subset consisting of 56,296 subjects (participating in at least two surveys), including 122 incident cases of autoimmune diabetes in adults (aged ≥35 and anti-GAD positive) and 1,555 cases of type 2 diabetes (aged ≥35 and anti-GAD negative). Hazard ratios (HRs) of diabetes associated with self-reported education and occupation were estimated by Cox proportional hazards models.RESULTSHigh levels of education (university versus primary school) were associated with an increased risk of autoimmune diabetes (HR 1.98 [95% CI 1.21–3.26]), after adjustment for BMI, lifestyle factors, and family history of diabetes. Case subjects with high levels of education had lower levels of C-peptide, tended to have higher levels of anti-GAD, and were more often treated with insulin. Conversely, these subjects had a reduced risk of type 2 diabetes (HR 0.69 [95% CI 0.57–0.82]), a risk that was partly explained by lower BMI and more physical activity (adjusted HR 0.89 [95% CI 0.74–1.06]).CONCLUSIONSHigh levels of education are associated with an increased risk of autoimmune diabetes in adults, a finding that may be mediated by effects on autoimmune activity. Because the association is not explained by traditional risk factors, other, currently unidentified, environmental factors are likely to be involved.